JOURNAL OF MORPHOLOGY 237:53–67 (1998)
Experimental Analysis of Character Coupling Across a Complex
Life Cycle: Pigment Pattern Metamorphosis in the Tiger
Salamander, Ambystoma tigrinum tigrinum
DAVID M. PARICHY*
Section of Evolution and Ecology and Center for Population Biology,
University of California at Davis, Davis, CA 95616
ABSTRACT Developmental relationships among characters are expected to
bias patterns of morphological variation at the population level. Studies of
character development thus can provide insights into processes of adaptation
and the evolutionary diversification of morphologies. Here I use experimental
manipulations to test whether larval and adult pigment patterns are coupled
across metamorphosis in the tiger salamander, Ambystoma tigrinum tigrinum (Ambystomatidae). Previous investigations showed that the early larval
pigment pattern depends on interactions between pigment cells and the
lateral line sensory system. In contrast, the results of this study demonstrate
that the major features of the adult pigment pattern develop largely independently of both the early larval pattern and the lateral lines. These results
suggest that ontogenetic changes that occur across metamorphosis decouple
larval and adult pigment patterns and could thereby facilitate independent
evolutionary modifications to the patterns during different stages of the life
cycle. J. Morphol. 237:53–67, 1998. r 1998 Wiley-Liss, Inc.
KEY WORDS: neural crest; lateral line; developmental constraint; complex life cycle; pigment
pattern evolution
Knowledge of the developmental relationships among characters is essential for a
complete understanding of morphological
variation. This is because the hierarchical
and interactive nature of development makes
some combinations of states across characters more likely than others. For example,
when several characters depend on the same
genes, mutations affecting those genes often
have pleiotropic effects (Wright, ’68). Similarly, when cells that give rise to different
characters interact during development, mutations that influence the behavior of one
cell population can have cascading, indirect
effects on the others (Atchley and Hall, ’91).
Such biases on the generation of variant
phenotypes (or ‘‘developmental constraints’’
[Maynard Smith et al., ’85]) can contribute
to genetic and phenotypic correlations at the
population level (Riedl, ’78; Cheverud, ’84,
’95; Riska, ’86; Kingsolver and Wiernasz,
’91; Arnold, ’92; Cowley and Atchley, ’92; but
see Gromko, ’95) and may influence the direction or rate of morphological evolution (Alberch, ’80; Oster et al., ’88; Wake, ’91; Shubin et al., ’95; also see Charlesworth et al.,
r 1998 WILEY-LISS, INC.
’82). For instance, selection on one character
may lead to the correlated evolution of other
developmentally related characters that are
expressed at the same stage of ontogeny
(Falconer, ’89; Riska, ’89; Price and Langen,
’92; Hanken and Wake, ’93; Price and
Pavelka, ’96).
Developmental relationships among characters also may affect the evolution of phenotypes across stages. Indeed, a character that
is expressed at different stages can be viewed
as a set of interdependent characters that
share some or all of their developmental
determinants (Atchley and Hall, ’91; Cowley
and Atchley, ’92). Selection favoring a particular character state during one stage thus
can result in the correlated evolution of the
same character or developmentally related
Contract grant sponsor: NSF; Contract grant number: IBN9423116; Contract grant number: IBN-9509802; Contract grant
sponsor: NIH; Contract grant number: GM53258.
*Correspondence to: David M. Parichy, Department of Genetics,
Box 8232, Washington University School of Medicine, 4566 Scott
Avenue, St. Louis, MO 63110. E-mail: [email protected]
54
D.M. PARICHY
characters at other stages, such that phenotypes expressed in adults may reflect selection at preadult stages or the reverse (Riska
and Atchley, ’85; Slatkin, ’87; Ebenman, ’92).
Alternatively, selection may favor developmentally incompatible character states
across stages. For example, Price and Grant
(’84) observed selection for small body size
among juvenile finches but selection for large
body size among adults (also see Clegg et al.,
’78; Roach, ’86; Kaplan, ’92; Chippindale et
al., ’96). In such instances, evolutionary
change may be slow at both stages, or adaptive modifications during one stage may be
accompanied by maladaptive, correlated responses at the other stage (Arnold, ’92; Ebenman, ’92; Price and Langen, ’92; Kirkpatrick
and Lofsvold, ’92; Chippindale et al., ’96).
A potential solution to the problem of differing and possibly conflicting selection
across stages is the evolution of a complex
life cycle, in which individuals undergo a
metamorphosis one or more times during
development. Complex life cycles and their
associated metamorphoses are present in
the majority of animals and are most often
viewed as a means for decoupling traits that
are expressed during different stages,
thereby permitting independent adaptations to different environments and selective
regimes (Haldane, ’32; Szarski, ’57; Istock,
’67; Wassersug, ’75; Wilbur, ’80; Werner, ’88;
Ebenman ’92; Moran, ’94)1. Support for this
hypothesis of decoupling across life cycle
stages comes from two principal sources.
First, closely related taxa often resemble
each other at one life cycle stage but differ
greatly at another (de Beer, ’58; Strathman,
’78; Williamson, ’82; Wray, ’92, ’96; also see
Wake, ’89; Wake and Hanken, ’96), suggesting that independent modifications have occurred within stages without affecting phenotypes across stages. Such observations
provide retrospective evidence that pre- and
postmetamorphic characters may be uncoupled from one another. Yet these studies
provide relatively little insight into the extent of developmental coupling across life
cycle stages for other characters in other
extant taxa, particularly since the degree of
morphological reorganization at metamor-
1Developmental coupling of trait expression is an alternative
and less frequently considered explanation for the evolutionary
persistence of complex life cycles in some phylogenetic lineages
(e.g., a larval stage might be maintained even in the face of
selection for its loss if larval morphology is a necessary precursor
to adult morphology [for a fuller discussion see Moran, ’94]).
phosis varies among phylogenetic lineages
(e.g., salamanders change relatively little at
metamorphosis compared to frogs [Wassersug and Hoff, ’82; Duellman and Trueb, ’86;
also see Shaffer and Lauder, ’88; Ashley et
al., ’91]).
Second, intraspecific descriptive studies
often reveal dramatic changes in morphology, physiology, and behavior at metamorphosis (Gilbert and Frieden, ’81; Balls and
Bownes, ’83; Gilbert et al., ’96). This suggests intuitively that pre- and postmetamorphic character states may be independent of
one another. For example, metamorphosing
amphibians undergo a host of changes in the
craniofacial skeleton, skin, pigmentation,
gastrointestinal tract, immune system, and
other characters (Wilder, ’25; Noble, ’31; Gilbert and Frieden, ’81; Fox, ’84; Duellman
and Trueb, ’86; Reilly and Lauder, ’90; Shaffer et al., ’91; Rose and Reiss, ’93; Hourdry et
al., ’96). Yet descriptive approaches often
cannot provide strong evidence for or against
developmental coupling. For instance, characters that look very different and thus might
appear superficially to be uncoupled may
nevertheless covary if they share common
underlying developmental genetic mechanisms (e.g., Besmer et al., ’93; Ingham, ’95;
van Eeden et al., ’96). Indeed, many of the
same genes involved in patterning during
embryogenesis are reexpressed during the
development of adult tissues, and results of
laboratory studies suggest that naturally
occurring mutations may affect both preand postmetamorphic characters (e.g., Shellenbarger and Mohler, ’78; Brabant and
Brower, ’93; Patterton et al., ’95; Ranganayakulu et al., ’95; Rauskolb et al., ’95; Stolow
and Shi, ’95).
More definitive evidence concerning developmental coupling across stages of complex
life cycles can be gained from quantitative
genetic or experimental approaches. These
strategies have been employed in studies of
life history characters (e.g., Blouin, ’92; Emlet and Hoegh-Guldberg, ’97), but they typically have not been used to assess the extent
of developmental coupling between discrete
pre- and postmetamorphic morphological
characters.
In the present study, I test experimentally
whether pigment patterns are coupled across
metamorphosis in the Eastern tiger salamander, Ambystoma tigrinum tigrinum
(family Ambystomatidae), a species in which
individuals hatch as aquatic larvae, grow
PIGMENT PATTERN METAMORPHOSIS
rapidly, and then metamorphose into terrestrial adults. Pigment patterns have been
used to define species and subspecies comprising the geographically wide-ranging Ambystoma tigrinum—complex of salamanders
(Dunn, ’40; Gehlbach, ’67; also see Shaffer,
’93; Shaffer and McKnight, ’96). These patterns are likely to be functionally important:
they are distinctive at hatching, when salamanders are especially vulnerable to predation (Stine et al., ’54; Anderson et al., ’71;
Kusano et al., ’85), and pigment patterns of
adults could be cryptic or aposematic depending on ecological context (Carpenter, ’55; Norris and Lowe, ’64; Hensel and Brodie, ’76;
Endler, ’78; Collins et al., ’80). Previous studies showed that a major element of the early
larval pattern depends on interactions between migrating pigment cells and the lateral lines (see below), a bilateral sensory
system that detects mechanical stimuli via
periodically arranged neuromasts and functions in orientation, feeding, and predator
avoidance (Wright, ’51; Atema et al., ’88;
Blaxter and Fuiman, ’90; Montgomery et al.,
’97). Here I use microsurgical manipulations
to test whether larval and adult pigment
patterns of A. t. tigrinum are coupled developmentally. If patterns are coupled across
life cycle stages, an analysis of the evolutionary origins and adaptive significance of the
adult pattern would have to consider not
only selection occurring after metamorphosis but also selection occurring in the aquatic,
larval environment.
DEVELOPMENTAL BACKGROUND
Ectothermic vertebrates possess three
principal types of pigment cells, or chromatophores: black melanophores, yellow xanthophores, and silvery iridophores (DuShane,
’43; Erickson, ’93; Frost-Mason et al., ’95).
All are derived from neural crest cells, which
arise along the dorsal neural tube shortly
after neurulation and then disperse throughout the embryo. Neural crest cells also contribute to the teeth, craniofacial skeleton,
peripheral nervous system, heart, endocrine
glands, fin mesenchyme, and other characters (Hall and Hörstadius, ’88; Selleck et al.,
’93).
Early larval pigment patterns
The pigment patterns of salamanders
shortly after hatching consist primarily of
melanophores and xanthophores. A prominent feature of the early larval pattern in
55
A. t. tigrinum (and many other salamanders
in Ambystomatidae and Salamandridae) is a
region over the lateral face of the myotomes
where xanthophores are abundant but melanophores typically are not found (Parichy,
’96a). Since melanophores occur further dorsally and ventrally, the overall pattern consists of a pale yellow horizontal stripe bordered by dark dorsal and ventral stripes. In
A. t. tigrinum, formation of the early larval
pattern begins with melanophores scattering evenly over the myotomes. Xanthophores are confined initially to aggregates
dorsal to the neural tube, but these cells
then disperse, and as they do so adjacent
melanophores recede short distances to form
a series of alternating vertical bars over the
dorsal flank (Fig. 1) (also see Olsson and
Löfberg, ’92). Simultaneously, the trunk midbody lateral line begins to develop through
the deployment of a migrating lateral line
‘‘primordium,’’ which causes melanophores
in its vicinity to retreat dorsally and ventrally, thereby establishing a subtle melanophore-free region. This region is then colonized by xanthophores and becomes
progressively more distinctive. The lateral
lines are responsible for initiating this pattern in A. t. tigrinum; a melanophore-free
region does not form when lateral line development is prevented. Similar effects of the
lateral lines have been demonstrated in several other ambystomatids and salamandrids
(Parichy, ’96b,c; for correlations in Plethodontidae and fishes, see Noble, ’31; Mabee, ’95).
Postmetamorphic pigment patterns
Pigment patterns of adult salamanders
typically are composed of melanophores, xanthophores, and iridophores. Adult A. t. tigrinum display a black ground color of dermal
melanophores and light spots of dermal xanthophores and iridophores (Bishop, ’41;
DuShane, ’43). Epidermal melanophores are
present but do not contribute to the overall
appearance of the pigment pattern. Little is
known of the developmental mechanisms
underlying adult pigment patterns in amphibians (Frost-Mason et al., ’95; Reedy et
al., ’98). Embryological grafting experiments
have suggested roles for embryonic ectoderm and mesoderm in promoting the development of different classes of adult pigment
cells in A. maculatum (Lehman, ’53), but
these experiments also may have perturbed
lateral line development, complicating the
interpretation of these results (see Parichy,
’96B). Nevertheless, these and other studies
56
D.M. PARICHY
Fig. 1. Formation of a melanophore-free region and
horizontal stripe pattern correlates with the development of the trunk midbody lateral line in Ambystoma
tigrinum tigrinum. The trunk lateral lines develop from
cranial ectodermal lateral line placodes that deploy migrating lateral line primordia. These primordia travel
caudally within the epidermis and deposit clusters of
mechanosensory cells at periodic intervals that later
erupt through the epidermis as mature, mechanoreceptive neuromasts. Three trunk lateral lines develop in
the sequence: midbody, dorsal, and ventral (Northcutt et
al., ’94). A: Brightfield micrograph showing the distribution of melanophores at stage 36/37 (Bordzilovskaya et
al., ’89) (prior to hatching) and the initially subtle melanophore-free region (arrow) as well as developing vertical bars of xanthophores (arrowheads). B: Corresponding fluorescence double exposure showing the migrating
midbody lateral line primordium (large arrow) and dorsal lateral line primordium (small arrow), as well as
xanthophores (autofluorescing cells) dispersing from premigratory aggregates (original positions indicated with
arrowheads). The lateral lines are labeled with a fluorescent vital dye (for details see Parichy, ’96b,c). In all photographs, anterior is to the right. Scale bar 5 500 µm.
(Smith-Gill, ’74; Bagnara, ’82) do suggest
that some features of adult patterns are
determined as early as the larval feeding
stage, implying that larval and adult patterns might share some developmental determinants. Consistent with the hypothesis that
pre- and postmetamorphic patterns are
coupled in A. t. tigrinum, several related
subspecies exhibit less distinctive melanophore-free regions in larvae and also display
more variable patterns in adults (Dunn, ’40;
Bishop, ’62; Stebbins, ’85; Parichy, ’96a, unpublished data). Likewise, the brightness of
the late larval pattern correlates with the
brightness of the postmetamorphic pattern
in A. t. nebulosum (Fernandez and Collins,
’88). Finally, iridophores are a major compo-
nent of the adult pattern, and some of the
first of these cells to differentiate in larvae
do so in association with lateral line neuromasts (see below), raising the possibility that
the larval lateral lines may act as a patterning cue for cells contributing to adult spots.
If larval and adult pigment patterns of A. t.
tigrinum are coupled developmentally, such
coupling could arise because 1) the adult
pattern depends on the distribution of pigment cells constituting the early larval pattern, 2) the adult pattern shares with the
early larval pattern a dependence on the
lateral lines, or 3) the adult pattern depends
on both the early larval pattern and the
lateral lines. Under any of these scenarios,
prevention of lateral line development should
eliminate the larval melanophore-free region and also should perturb the adult pigment pattern.
MATERIALS AND METHODS
Embryos and manipulations
Ambystoma t. tigrinum embryos from a
natural population (supplied by the Charles
Sullivan Company, Nashville, TN) were
maintained in 20% Hepes-buffered Steinberg’s solution (HSS) (plus 37.5 IU/ml penicillin, 37.5 µg/ml streptomycin [Asashima et
al., ’89]) at 9–18°C. For microsurgeries, embryos representing five clutches were decapsulated with forceps, passed through four to
five changes of sterile 100% HSS (plus 75
IU/ml penicillin, 75 µg/ml streptomycin), and
then placed into sterile agar-lined dishes
containing 100% HSS. To prevent the development of trunk lateral line primordia,
nerves, and neuromasts, ectoderm containing the lateral line placodes (Northcutt et
al., ’94) was removed unilaterally with tungsten needles and then replaced with a patch
of donor belly epidermis (Fig. 2) (Parichy,
’96b,c). Manipulations were performed on
embryos at stages 28–31 (Bordzilovskaya et
al., ’89), which is prior to the onset of dispersal for most or all trunk neural crest cells
(Olsson and Löfberg, ’92). A series of embryos were sham-manipulated by removing
and then replacing placode-area ectoderm
unilaterally. For both lateral line–ablated
and sham-manipulated embryos, operated
sides were chosen at random. To test for
nonspecific surgical effects, additional embryos were treated identically but were left
unmanipulated.
57
PIGMENT PATTERN METAMORPHOSIS
Fig. 2. Microsurgical procedure for ablating the trunk
lateral lines in Ambystoma tigrinum tigrinum. To prevent lateral line development unilaterally, placode-area
ectoderm was removed and replaced with belly epidermis from a similarly staged donor.
Rearing conditions
Salamanders were maintained individually (15°C; 14L:10D) throughout the experiment. Embryos and early larvae were reared
in 60 mm Petri dishes containing 20% HSS.
Shortly after the onset of feeding, larvae
were transferred to plastic dishes (100 3 40
mm) containing 50% Holtfreter’s solution
(Asashima et al., ’89). As individuals grew,
they were transferred to plastic boxes (30 3
15 3 9 cm) containing aged tap water. Postmetamorphic salamanders were housed in
plastic boxes lined with moist foam pads.
Containers were rearranged at each cleaning to minimize position effects: dishes were
rearranged daily; boxes containing larvae
were rearranged every third day; boxes containing metamorphosed salamanders were
rearranged once per week. Early larvae were
fed newly hatched brine shrimp twice daily
and then were acclimated to a diet of tubifex
worms once per day. As larvae approached
metamorphosis, they were fed tubifex worms
and crickets. Metamorphosed salamanders
were fed crickets dusted with calcium and
vitamins three times per week. Days of development were calculated relative to completion of the early larval pattern (stage 41
[Parichy, ’96b]); salamanders were considered metamorphosed when their gills and
tail fin had regressed almost entirely and
they could be transferred to foam pads.
Quantitative methods
Characterization of early larval pigment
patterns has been described (Parichy, ’96b).
To document postmetamorphic patterns,
salamanders were first anesthetized and
then cradled in molds of modeling clay within
a water-filled basin. A sheet of glass was
rested on spacers and pressed against the
salamander to flatten the upper surface of
the integument. Images of left and right
sides were captured with a Sony CCD videocamera and macro lens interfaced to an Apple
Macintosh computer running the public domain NIH Image program (Wayne Rasband,
NIH). Digital images were transferred to
Adobe Photoshop, and a rectangular region
between the second and ninth costal grooves
and the dorsal and ventral margins of the
flank was analyzed. Light spots of xanthophores and iridophores were traced manually and normalized to white; all dark ground
color was normalized to black. Dorsoventral
image heights were normalized to 100 pixels.
To describe patterns, I calculated the proportional total area covered by spots for each
image as the number of white pixels divided
by the total number of pixels. Proportional
regional areas covered by spots were similarly quantified for five equally sized positions at different dorsoventral levels of the
flank. The number of positions was chosen a
priori based on examination of adult patterns on unmanipulated sides (without regard to patterns on manipulated sides) and
represented a compromise between maximizing resolution and minimizing the impact of
minor variation in the positioning of salamanders. A posteriori comparisons using different total numbers of positions yielded
qualitatively similar results (data not
shown). Also calculated were the numbers of
spots and the mean perimeters and mean
areas of spots (in pixels). Finally, the mean
perimeter:area ratios of spots were calculated as a measure of shape since these
ratios increase as spots of a given area depart from circularity. Differences between
unmanipulated and manipulated sides (or
left and right sides of unmanipulated controls) were assessed with paired t-tests, using arcsine transformations for ratio data
(Sokal and Rohlf, ’81). This allowed testing
for effects on pigment patterns after controlling for interindividual variability.
RESULTS
To test whether early larval and adult
pigment patterns are coupled developmentally, I prevented lateral line development
unilaterally and reared the experimentally
58
D.M. PARICHY
manipulated and control embryos (n 5 131)
through metamorphosis.
Early larval pigment patterns
On unmanipulated sides of early larvae
(stage 41, shortly after hatching normally
would occur; approximately 15 mm total
length), melanophores were found over the
dorsal myotomes and further ventrally at
the dorsal margin of the yolk mass, but
relatively few melanophores were found over
the lateral face of the myotomes. On lateral
line–ablated sides, however, melanophores
more extensively colonized the flank (n 5
84) (Fig. 3). Figure 4 presents a reanalysis of
melanophore distributions (from Parichy,
’96b), in which melanophore densities are
pooled within four dorsoventral regions of
the flank (rather than the original 15 regions), to facilitate comparison with analyses of postmetamorphic patterns (below). On
lateral line–ablated sides (Fig. 4A), melanophore densities were lower near the dorsal
edge of the myotomes (0–400 µm from the
base of the dorsal fin; paired t42 5 6.84, P ,
0.0001) and higher over the lateral face of
the myotomes (401–800 µm; paired t42 5
12.39, P , 0.0001), in agreement with Par-
Fig. 4. Melanophore distributions in early larvae depend on the lateral lines in Ambystoma tigrinum tigrinum.
Presented are mean melanophore densities (695% confidence intervals) from reanalyses of lateral line–ablated (A)
and sham-manipulated (B) larvae of Parichy (’96b). Positions represent distances from the base of the dorsal fin and
the dorsal apex of the myotomes, and only midpoints are
shown. Filled bars, unmanipulated sides; open bars, manipulated sides. A: Lateral line ablation (43 larvae; 14,489 melanophores) resulted in more uniform distributions of melanophores, with lower densities dorsally and higher densities
over the middle of the myotomes on lateral line–ablated
sides (open bars) as compared to lateral line–intact sides
(filled bars). B: Sham manipulation (15 larvae; 4,021 melanophores) typically did not affect melanophore densities on
manipulated sides (open bars) as compared to unmanipulated sides (filled bars), though a subtle increase in melanophore density could sometimes be observed in the middle of
the flank (401–800 µM) (see Parichy, ’96b) probably due to
minor damage inflicted on the lateral line placodes during
removal and replacement. *Significantly different at a 5
0.05 level with sequential Bonferonni correction for four
comparisons (Rice, ’89).
Fig. 3. Prevention of lateral line development eliminates the melanophore-free region and horizontal stripe
pattern in early larval Ambystoma tigrinum tigrinum.
Opposite sides of a single individual (stage 41) with the
image in B flipped to facilitate comparison with A. A: On
the unmanipulated side, a distinctive melanophore-free
region is found over the lateral face of the myotomes at
the level of the midbody lateral line (large arrow). Arrowheads indicate the positions of xanthophore bars. B: On
the lateral line–ablated side, melanophores readily colonize this area. Scale bar 5 1 mm.
ichy (’96b). Ablation of the lateral lines also
resulted in a statistically significant 7% increase in total melanophore density and perturbed the distribution of xanthophores,
which remained principally between vertical bars of melanophores (see Parichy, ’96b,c).
PIGMENT PATTERN METAMORPHOSIS
Sham manipulations (n 5 27) did not affect
melanophore distributions (Fig. 4B).
59
Later larval pigment patterns
No differences in behavior, growth, or vigor
were observed among lateral line–ablated,
sham-manipulated, and unmanipulated larvae. By the middle of the larval period (approximately 70 mm total length) in salamanders of all treatments, melanophores
were apparent in dorsal regions of the flank
previously occupied only by xanthophores
(Fig. 5A). Nevertheless, a distinctive melanophore-free region was still evident over the
lateral face of the myotomes on unmanipulated sides, centered on the midbody lateral
line. Iridophores were present subjacent to
midbody and ventral lateral line neuromasts, lining the peritoneum, and were also
scattered irregularly over the trunk. On lat-
eral line–ablated sides, a melanophore-free
region was not apparent (Fig. 5C).
At later larval stages (approximately 100
mm total length), the pattern gradually
transformed to a mottled green and black in
salamanders of all treatments (Fig. 5B). This
change in coloration may reflect the invasion of the dermis by melanophores previously localized beneath the subepidermal
basement membrane, which thickens gradually and is occupied by mesenchymal cells to
form the definitive dermis (Stearner, ’46). A
distinctive melanophore-free region was no
longer apparent over the middle of the myotomes on lateral line–intact sides. Nevertheless, a region of lighter pigmentation persisted in the vicinity of the midbody lateral
line, and small spots of iridophores delineated the positions of neuromasts. These
Fig. 5. Lateral line effects on the pigment pattern
persist through the larval period of Ambystoma tigrinum tigrinum. Shown are opposite sides of the same
larva presented in Figure 3, at day 47 (A,C) and day 95
(B,D). A: On the unmanipulated side, a distinctive melanophore-free region is still present in the vicinity of the
midbody lateral line (large arrow) during the middle
larval period, and the positions of lateral line neuromasts are delineated by concentrations of iridophores
(one is indicated at the tip of the large arrow). Arrowheads indicate regions now occupied by melanophores
that correspond to the positions of xanthophore bars at
earlier stages. Small arrow, level of the ventral lateral
line. B: During the late larval period, a region of lighter
pigmentation is still centered on the midbody lateral
line (large arrow), but the distinctiveness of this region
is considerably diminished. Small arrow, level of the
ventral lateral line. C: On the lateral line–ablated side,
melanophores are more uniformly distributed during
the middle larval period as compared to the unmanipulated side in A. D: The consequences of lateral line
ablation are still manifested during the late larval period, though the magnitude of the difference compared
to the lateral line–intact side in B is reduced. Images in
C and D are reversed to facilitate comparison with A and
B, respectively. Scale bars 5 5 mm.
60
D.M. PARICHY
features were not present on lateral line–
ablated sides (Fig. 5D).
Postmetamorphic pigment patterns
Every salamander (n 5 131) survived for
scoring of the adult pigment pattern approximately 42 weeks after development of the
early larval pattern and on average 29 weeks
after metamorphosis. Times to metamorphosis did not differ among lateral line–ablated,
sham-manipulated, and unmanipulated
salamanders (pooled mean 5 99 d, SD 5 7.1,
range 5 80–119; F2,128 5 0.15, P 5 0.9).
Upon completion of the experiment, treatment groups did not differ in snout-to-vent
lengths (pooled mean 5 103 mm, SD 5 4.1,
range 5 97–120; F2,128 5 0.37, P 5 0.7) or
total lengths (pooled mean 5 204 mm, SD 5
11.8, range 5 190–243; F2,128 5 0.64, P 5
0.5). Thus, manipulations did not adversely
affect general growth and development under these conditions.
Postmetamorphic salamanders displayed
pigment patterns typical of A. t. tigrinum: a
black ground color with dark yellow-brown
spots dorsally and lighter yellow spots ventrally. After pooling unmanipulated sides of
lateral line–ablated and sham-manipulated
salamanders, and arbitrarily chosen sides of
unmanipulated controls total proportional
areas covered by spots did not differ among
clutches (F4,126 5 1.63, P 5 0.2) and were not
correlated with times to metamorphosis (r 5
0.03, P 5 0.8) but tended to increase with
body size (correlations with snout-to-vent
length and total length: r 5 0.28, 0.30; P ,
0.05). At metamorphosis, neuromasts are
covered by epidermis and presumably regress to a less-differentiated state as in other
salamanders (Noble, ’31; Dawson, ’36;
Wright, ’51; Fritzsch et al., ’88); well-differentiated neuromasts were not observed in histological sections of integument at the end of
the experiment (not shown).
Effects of lateral line ablation on early
larval pigment patterns (Figs. 3, 4) suggested the hypothesis that correlated effects
in adults might be manifested at dorsolateral levels of the flank. Nevertheless, no
major effects of lateral line ablation on the
adult pattern were evident (n 5 84) (Fig. 6),
and comparisons of areas covered by spots at
corresponding dorsoventral positions did not
reveal effects of lateral line ablation in the
most dorsal four of five regions (Fig. 7A). In
the most ventral region, however, lateral
line–ablated sides had slightly but significantly less area covered by spots (paired
t83 5 4.54, P , 0.0001), and these spots
tended not to form as orderly or continuous a
row as compared to lateral line–intact sides
(Fig. 6A,C). This region corresponds to the
position of the ventral lateral line in larvae;
most melanophores and xanthophores constituting the early larval pattern are not localized this far ventrally, and larval melanophore density is not perturbed at this level
following lateral line ablation (Figs. 3, 4A, 5)
(Parichy, ’96b). Total areas covered by spots
across the entire flank did not differ significantly due to lateral line ablation (1% less on
lateral line–ablated sides; paired t83 5 1.80,
P 5 0.08). Among control salamanders, total
and regional areas covered by spots did not
differ between unmanipulated and shammanipulated sides (total areas: paired t26 5
0.52, P 5 0.6, n 5 27; regional areas: see Fig.
7B) or between left and right sides of unmanipulated individuals (total areas: paired
t19 5 0.47, P 5 0.6, n 5 19; regional areas:
data not shown).
Spot numbers as well as mean spot areas,
perimeters, and perimeter-to-area ratios did
not differ between lateral line–ablated and
lateral line–intact sides (mean perimeters:
8% less on lateral line–ablated sides; paired
t83 5 1.79, P 5 0.08; all others: P . 0.2) (data
not shown). Visual inspections by observers
blind with respect to treatment failed to
identify other lateral line effects.
DISCUSSION
Interactions between pigment cells and
the lateral line sensory system contribute to
a distinctive horizontal stripe pattern in larval salamanders. This study demonstrates
that in A. t. tigrinum lateral line effects on
chromatophores persist through middle larval stages but diminish as metamorphosis
approaches, and the postmetamorphic pigment pattern arises largely independently
of both the lateral lines and the distribution
of pigment cells that contribute to larval
stripes. These findings are consistent with a
model in which ontogenetic changes that
occur across metamorphosis decouple larval
and adult pigment patterns, and could
thereby facilitate independent evolutionary
modifications to the patterns during different phases of the life history.
Pigment pattern development
Prevention of lateral line development profoundly altered the distributions of melanophores and xanthophores over the dorsal
PIGMENT PATTERN METAMORPHOSIS
61
Fig. 6. Major features of the adult pigment pattern
do not depend on the lateral lines or the distribution of
pigment cells comprising the early larval pattern in
Ambystoma tigrinum tigrinum. Shown is the same individual presented in Figs. 3 and 5 after metamorphosis
on day 299. A: On the unoperated side, light yellow and
brown spots are typical of the normal adult pigment
pattern in A. t. tigrinum. B: Dorsal view of the same
salamander. a, lateral line–ablated side; i, lateral line–
intact side. C: On the side without lateral lines, spots in
dorsal and lateral regions of the flank (corresponding to
the position of the midbody lateral line in larvae) are not
obviously perturbed, though bright spots along the ventrolateral flank (arrow) are somewhat less regular and
continuous as compared to the unmanipulated side in A.
This image is reversed to facilitate comparison with A.
D: Ventral view. Scale bar 5 10 mm.
and lateral flank in early larvae (also see
Parichy, ’96b,c) but did not yield pigment
pattern defects in the corresponding regions
of adults. This failure to detect significant
effects of lateral line ablation in dorsolateral
regions of the flank after metamorphosis
probably is not due to insufficient statistical
power. Estimates of error were low (see Fig.
7), and even a subtle effect of the lateral
lines was detectable ventrolaterally, where
lateral line–ablated sides had a 3% deficit in
the area covered by spots (presumably contributing to marginal effects of lateral line
ablation on total areas covered by spots and
mean spot perimeters). Since early larval
patterns in this region do not depend on the
lateral lines, this difference could indicate
that the ventral lateral line has a direct
(albeit minor) influence on cells that contribute to adult ventral spots, rather than acting
indirectly via some effect on early larval
chromatophores. For example, the ventral
lateral line might stimulate the proliferation or differentiation of dermal iridophores
that appear during later larval stages or
might act as a patterning cue that contributes to the localization of these cells. The
present observations also cannot exclude the
possibility that larval and adult patterns
might be coupled through other shared patterning mechanisms that have yet to be identified. For instance, chromatophores of both
larvae and adults are derived from neural
crest cells (Twitty and Bodenstein, ’39;
DuShane, ’43), so mutations affecting general properties of these cells (e.g., morphogenetic behaviors or specification of chromatophore lineages) might yield pattern defects
at both stages. Despite these caveats, however, the general conclusion suggested by
this study is that the major features of the
adult pigment pattern in A. t. tigrinum do
not depend on either the distribution of pigment cells contributing to larval stripes, or
the midbody lateral line, despite the latter’s
pivotal role in establishing the early larval
pattern.
62
D.M. PARICHY
Fig. 7. Lateral line ablation has only subtle effects
on the adult pigment pattern in Ambystoma tigrinum
tigrinum. Mean areas covered by spots at different dorsoventral positions of the flank are presented for lateral
line–ablated (A) and sham-manipulated (B) individuals.
Filled bars, unmanipulated sides; open bars, manipulated sides. A: Areas covered by spots did not differ
between lateral line–intact sides (filled bars) and lateral
line–ablated sides (open bars) at dorsal and lateral
regions (positions 1–4) but differed significantly in the
most ventral region (position 5). B: Areas covered by
spots did not differ between unmanipulated (filled bars)
and sham-manipulated sides (open bars). Error bars are
95% confidence intervals converted back to ratio scale
after arcsine transformation. *Significantly different at
a 5 0.05 level with sequential Bonferonni correction for
five comparisons (Rice, ’89).
The relative independence of larval and
adult pigment patterns in A. t. tigrinum
could be achieved through either of two developmental strategies (also see Alberch, ’87;
Moran, ’94). If the same population of melanophores and xanthophores in larvae gives
rise to the pattern in adults, then morphogenetic remodeling of this population through
differential proliferation or spatial rearrangements could contribute to decoupling
larval and adult phenotypes. Several studies are consistent with these possibilities
(Berweger, ’26; Woronzowa, ’32; Stearner,
’46; Lehman, ’53; Yasutomi, ’87; Lechaire
and Denefle, ’91), though direct evidence is
lacking. Or, different populations of pigment
cells could contribute to patterns across
stages. For example, chromatophores constituting the adult pattern might arise from
precursor cells that are set aside at embyronic stages and recruited to differentiate
only at metamorphosis. Consistent with such
compartmentalization of larval and adult
cell lineages, adult melanophores differentiate de novo at metamorphosis in the newt
Taricha torosa (family Salamandridae) (see
below), and distinct populations of melanophores contribute to larval and adult pigment patterns in the zebrafish, Danio rerio
(Johnson et al., ’95). Similarly, the adult
epibranchial cartilage of the salamander Eurycea bislineata arises from just a few cells
in the perichondrium of the larval, neural
crest–derived epibranchial cartilage (Alberch and Gale, ’86), and different populations of cells contribute to larval and adult
epidermis, intestine, and musculature in anuran amphibians (Alley, ’89; Kinoshita and
Sasaki, ’94; Hourdry et al., ’96; Furlow et al.,
’97). Nevertheless, the presence of ‘‘latent’’
pigment cell precursors in the skin of A. t.
tigrinum has yet to be demonstrated. If such
a population does exist, the results of this
study indicate that it must be under sufficiently different control so as to be unaffected by defects in the arrangements of
larval chromatophores.
Whichever developmental strategy or combination of strategies is responsible for the
transition from a larval to adult pigment
pattern in A. t. tigrinum, the extent of this
transformation can be contrasted with other
taxa. For example, Taricha torosa larvae
exhibit horizontal stripes of melanophores,
but, unlike other ambystomatids and salamandrids, stripes in T. torosa do not depend
on the lateral lines because redundant, evolutionarily derived patterning mechanisms
have been layered over the primitive lateral
line–dependent mechanisms (Parichy, ’96b,
in preparation). This species also may exhibit derived mechanisms for the development of the uniform orange-brown adult pattern since, unlike Ambystoma, dermal
melanophores that constitute the larval pattern degenerate at metamorphisis and are
replaced by a second population of melanophores that differentiates de novo in the
epidermis (Stearner, ’46; Niu and Twitty, ’50;
McCurdy, ’71). Thus, mechanisms under-
PIGMENT PATTERN METAMORPHOSIS
lying larval pattern formation appear to be
more complex, and the larval-to-adult transformation appears to be more extensive in T.
torosa than in A. t. tigrinum. At a deeper
phylogenetic level, pigment pattern metamorphosis in many frogs involves the organization of chromatophores into functionally
and physiologically integrated dermal chromatophore units (Bagnara et al., ’68), in
contrast to the much simpler arrangements
of chromatophores in adult salamanders
(Stearner, ’46). Increased complexity of patterning mechanisms within stages and
greater disparity of mechanisms across
stages (i.e., increased pattern modularity)
may reflect conflicting selection in larval
and adult environments (Riska, ’86; Raff,
’96; Wagner and Altenberg, ’96), although
the fitness consequences of variation in morphological characters expressed both before
and after metamorphosis remain largely unexplored.
Consequences of metamorphosis
What are the consequences of metamorphosis for the development and evolution of
pigment patterns and other characters?
Complex life cycles and their associated
metamorphoses are generally thought to decouple characters that are expressed at different stages and in different selective regimes, thereby allowing stage-specific
adaptations for resource acquisition, dispersal, reproduction, or other activities (Istock, ’67; Bryant, ’69; Wassersug, ’75; Wilbur,
’80; Werner, ’88; Ebenman, ’92). This ‘‘adaptive decoupling’’ hypothesis (Moran, ’94) predicts that traits expressed before and after
metamorphosis should depend on different
developmental mechanisms and gene activities (Haldane, ’32; Alberch, ’87; Ebenman,
’92). The finding that the adult pattern of
A. t. tigrinum arises largely independently
of the lateral lines and the larval pattern is
thus consistent with the adaptive decoupling hypothesis and a role for metamorphosis in dissociating trait expression across life
cycle stages. Moreover, the same mechanisms that govern other events at metamorphosis also may regulate the transition from
the larval to the adult pigment pattern. Thyroid hormone (TH) triggers metamorphic
changes in a variety of characters (Gilbert
and Frieden, ’81; Rose and Reiss, ’93; Gilbert
et al., ’96), and results of several studies
suggest a possible role for TH in pigment
pattern transformation via direct effects on
chromatophore motility, proliferation, or dif-
63
ferentiation, or indirect effects on surrounding tissues (Woronzowa, ’32; Bagnara et al.,
’79; Yasutomi, ’87; Frost-Mason et al., ’95;
Brown, ’97). Nevertheless, the results of this
study do not test directly the role of metamorphosis per se in decoupling larval and adult
patterns. Indeed, characters may exhibit reduced interdependencies as more disparate
stages are compared even in taxa with simple
life cycles (Arnold, ’92; Kirkpatrick and Lofsvold, ’92; Cheverud et al., ’96). For example,
this study showed that midbody lateral line
effects on chromatophores were very distinctive during the first half of the larval period,
but were less apparent during later larval
development and were not detectable after
metamorphosis. Thus, alterations in the pigment pattern were not confined to the period
of metamorphic climax when other remodeling was most apparent (also see Bishop, ’41;
Stearner, ’46; Lehman, ’53), raising the possibility that pattern changes might not be
regulated by the same mechanisms that govern other metamorphic transformations. If
so, a TH-dependent metamorphosis might
not itself be responsible for decoupling patterns across life cycle stages. Consistent with
the idea that pattern transformation does
not depend on a metamorphic climax, several subspecies of A. tigrinum and the closely
related A. rosaceum, A. andersoni, and A.
taylori can develop adult markings even as
larvae or paedomorphs (Taylor, ’41; Anderson, ’61; Shaffer and McKnight, ’96; Shaffer
and Voss, personal communication). Alternatively, these observations could simply indicate that pigment patterns (or their determinants) differ from other characters in their
sensitivity to TH (e.g., Hanken et al., ’89;
Rose and Reiss, ’93; Rose, ’95). Experimental
perturbation of TH levels should illuminate
the extent to which pigment pattern transformation is developmentally and hormonally
integrated with other events at metamorphosis (Shaffer and Voss, ’96), and would provide a more direct test of the adaptive decoupling hypothesis for the role of metamorphosis
in dissociating the expression of pigment patterns and other characters (e.g., Shaffer et al.,
’91; Berger-Bishop and Harris, ’96; Vaglia et al.,
’97) across life cycle stages.
Manipulative experiments for studying
character evolution
Experimental studies of the mechanisms
underlying character development can
complement studies of character evolution
64
D.M. PARICHY
at the population level. Although manipulative approaches such as the one presented
here are not likely to provide insights into
subtle effects of alleles segregating in natural populations, they should reveal major
developmental interdependencies within and
among characters and can suggest predictions regarding morphological evolution. For
example, these data suggest that selection
on features of the lateral line sensory system
(such as receptor number) might yield correlated responses in larval but not adult patterns. Conversely, selection on the adult pattern would seem unlikely to result in the
correlated evolution of either the larval pattern or the lateral lines. These hypotheses
could be further tested by estimating quantitative genetic correlations across stages or
measuring correlated responses to artificial
selection (but see Gromko, ’95). More generally, a mechanistic understanding of character development will be essential for assessing the biological bases for character
correlations as well as long-term constraints
on their evolution (Riska, ’89; Arnold, ’92;
Shaw et al., ’95).
ACKNOWLEDGMENTS
I thank C.A. Erickson, R.K. Grosberg, J.
Hanken, F.W. Harrison, M.W. Hart, S.L.
Johnson, H.B. Shaffer, S.R. Voss, and one
anonymous reviewer for helpful discussions,
comments on the manuscript, or both. This
research has been supported by NSF dissertation improvement grant IBN-9423116 as
well as NIH GM53258 to C.A. Erickson and
NSF IBN-9509802 to H.B. Shaffer. I have
been supported by a NSF predoctoral fellowship, a fellowship from the Center for Population Biology (UC Davis), and scholarships
from the Northern California Regional Association of Phi Beta Kappa and the ARCS
Foundation (San Francisco).
LITERATURE CITED
Alberch, P. (1980) Ontogenesis and morphological diversification. Am. Zool. 20:653–667.
Alberch, P. (1987) Evolution of a developmental process:
Irreversibility and redundancy in amphibian metamorphosis. In R.A. Raff and E.C. Raff (eds): Development
as an Evolutionary Process. New York: Alan R. Liss,
pp. 23–46.
Alberch, P., and E.A. Gale (1986) Pathways of cytodifferentiation during metamorphosis of the epibranchial
cartilage in the salamander Eurycea bislineata. Dev.
Biol. 117:233–244.
Alley, K.E. (1989) Myofiber turnover is used to retrofit
frog law muscles during metamorphosis. Am. J. Anat.
184:1–12.
Anderson, J.D. (1961) The life history and systematics of
Ambystoma rosaceum. Copeia 1961:371–377.
Anderson, J.D., D.D. Hassinger, and G.H. Dalrymple
(1971) Natural mortality of eggs and larvae of Ambystoma t. tigrinum. Ecology 52:1107–1112.
Arnold, S.J. (1992) Constraints on phenotypic evolution.
Am. Nat. 140:S85–S107.
Asashima, M., G.M. Malacinski, and S.C. Smith (1989)
Surgical manipulation of embryos. In J.B. Armstrong
and G.M. Malacinski (eds): Developmental Biology of
the Axolotl. New York: Oxford University Press, pp.
253–263.
Ashley, M.A., S.M. Reilly, and G.V. Lauder (1991) Ontogenetic scaling of hindlimb muscles across metamorphosis in the tiger salamander, Ambystoma tigrinum.
Copeia 1991:767–776.
Atchley, W.R., and B.K. Hall (1991) A model for development and evolution of complex morphological structures. Biol. Rev. 66:101–157.
Atema, J., R.R. Fay; A.N. Popper, and W.N. Tavolga (eds)
(1988) Sensory Biology of Aquatic Animals. New York:
Springer-Verlag.
Bagnara, J.T. (1982) Development of the spot pattern in
the leopard frog. J. Exp. Zool. 224:283–287.
Bagnara, J.T., J.D. Taylor, and M.E. Hadley (1968) The
dermal chromatophore unit. J. Cell. Biol. 38:67–79.
Bagnara, J.T., J. Matsumoto, W. Ferris, S.K. Frost, W.A.
Turner, Jr., T.T. Tchen, and J.D. Taylor (1979) Common origin of pigment cells. Science 203:410–415.
Balls, M., and M. Bownes (eds) (1983) Metamorphosis.
New York: Oxford University Press.
Berger-Bishop, L.E., and R.N. Harris (1996) A study of
caudal allometry in the salamander Hemidactylium
scutatum (Caudata: Plethodontidae). Herpetologica 52:
515–525.
Berweger, L. (1926) Die Entwicklung der pigmentführenden Zellen in der Haut von Salamandra. Z. Mikrosk.
Anat. Forsch. 7:231–294.
Besmer, P., K. Manova, R. Duttlinger, E.J. Huang, A.
Packer, C. Gyssler, and R.F. Bachvarova (1993) The
kit-ligand (steel factor) and its receptor c-kit/W: pleiotropic roles in gametogenesis and melanogenesis. Development 1993(Suppl.):125–137.
Bishop, S.C. (1941) The salamanders of New York. N. Y.
State Mus. Bull. 324:1–365.
Bishop, S.C. (1962) Handbook of Salamanders. New
York: Hafner Publishing Company.
Blaxter, J.H.S., and L.A. Fuiman (1990) The role of the
sensory systems of herring larvae in evading predatory fishes. J. Mar. Biol. Assoc. U.K. 70:413–427.
Blouin, M.S. (1992) Genetic correlations among morphometric traits and rates of growth and differentiation in
the green tree frog, Hyla cinerea. Evolution 46:735–
744.
Bordzilovskaya, N.P., T.A. Detlaff, S.T. Duhon, and G.M.
Malacinski (1989) Developmental-stage series of axolotl embryos. In J.B. Armstrong and G.M. Malacinski
(eds): Developmental Biology of the Axolotl. New York:
Oxford University Press, pp. 201–219.
Brabant, M.C., and D.L. Brower (1993) PS2 integrin
requirements in Drosophila embryo and wing morphogenesis. Dev. Biol. 157:49–59.
Brown, D.D. (1997) The role of thyroid hormone in
zebrafish and axolotl development. Proc. Natl. Acad.
Sci. U. S. A. 94:13011–13016.
Bryant, E.H. (1969) A system favoring the evolution of
holometabolous development. Ann. Entomol. Soc.
Amer. 62:1087–1091.
Carpenter, C.C. (1955) Aposematic behavior in the salamander Ambystoma tigrinum melanostictum. Copeia
1955:311.
PIGMENT PATTERN METAMORPHOSIS
Charlesworth, B., R. Lande, and M. Slatkin (1982) A
neo-Darwinian commentary on macroevolution. Evolution 36:474–498.
Cheverud, J.M. (1984) Quantitative genetics and developmental constraints on evolution by selection. J.
Theor. Biol. 110:155–171.
Cheverud, J.M. (1995) Morphological integration in the
saddle-back tamarin (Saguinus fuscicollis) cranium.
Am. Nat. 145:63–89.
Cheverud, J.M., E.J. Routman, F.A.M. Duarte, B. van
Swinderen, K. Cothran, and C. Perel (1996) Quantitative trait loci for murine growth. Genetics 142:1305–
1319.
Chippindale, A.K., T.J. Chu, and M.R. Rose (1996) Complex trade-offs and the evolution of starvation resistance in Drosophila melanogaster. Evolution 50:753–
766.
Clegg, M.T., A.L. Kahler, and R.W. Allard (1978) Estimation of life cycle components of selection in an experimental plant population. Genetics 89:765–792.
Collins, J.P., J.B. Mitton, and B.A. Pierce (1980) Ambystoma tigrinum: A multispecies conglomerate? Copeia
1980:938–941.
Cowley, D.E., and W.R. Atchley (1992) Quantitative genetic models for development, epigenetic selection,
and phenotypic evolution. Evolution 46:495–518.
Dawson, A.B. (1936) Changes in the lateral line-organs
during the life of the newt, Triturus viridescens. A
consideration of the endocrine factors involved in the
maintenance of differentiation. J. Exp. Zool. 74:2221–
237.
de Beer, G. (1958) Embryos and Ancestors. New York:
Oxford University Press.
Duellman, W.E., and L. Trueb (1986) Biology of Amphibians. New York: McGraw-Hill.
Dunn, E.R. (1940) The races of Ambystoma tigrinum.
Copeia 1940:154–162.
DuShane, G.P. (1943) The embryology of vertebrate pigment cells. Part I: Amphibia. Q. Rev. Biol. 18:109–127.
Ebenman, B. (1992) Evolution in organisms that change
their niches during the life cycle. Am. Nat. 139:990–
1021.
Emlet, R.B., and O. Hoegh-Guldberg (1997) Effects of
egg size on postlarval performance: Experimental evidence from a sea urchin. Evolution 51:141–152.
Endler, J.A. (1978) A predator’s view of animal color
patterns. Evol. Biol. 11:319–364.
Erickson, C.A. (1993) From the crest to the periphery:
Control of pigment cell migration and lineage segregation. Pigment Cell Res. 6:336–347.
Falconer, D.S. (1989) Introduction to Quantitative Genetics. New York: Wiley.
Fernandez, P.J., Jr., and J.P. Collins (1988) Effect of
environment and ontogeny on color pattern variation
in Arizona tiger salamanders (Ambystoma tigrinum
nebulosum Hallowell). Copeia 1988:928–938.
Fox, H. (1984) Amphibian Morphogenesis. Clifton, NJ:
Humana Press.
Fritzsch, B., U. Wahnshaffe, and U.C. Bartsch (1988)
Metamorphic changes in the octavolateralis system of
amphibians. In B. Fritzsch, M.J. Ryan, W. Wilczynski,
T.E. Hetherington, and W. Walkowiak (eds): The Evolution of the Amphibian Auditory System. New York:
Wiley, pp. 359–376.
Frost-Mason, S.K., R. Morrison, and K.A. Mason (1995)
Pigmentation. In H. Heatwole and G. Barthalmus
(eds): Amphibian Biology. Sydney, Australia: Surrey
Betty & Sons, pp. 64–97.
65
Furlow, J.D., D.L. Berry, Z. Wang, and D.D. Brown
(1997) A set of novel tadpole specific genes expressed
only in the epidermis are down-regulated by thyroid
hormone during Xenopus laevis metamorphosis. Dev.
Biol. 182:284–298.
Gehlbach, F.R. (1967) Ambystoma tigrinum (Green): Tiger salamander. Catalogue of American Amphibians
and Reptiles 52:1–52.4.
Gilbert, L.I., and E. Frieden (eds) (1981) Metamorphosis: A Problem in Developmental Biology, 2nd ed. New
York: Plenum Press.
Gilbert, L.I., J.R. Tata, and B.G. Atkinson (eds) (1996)
Metamorphosis: Postembryonic Reprogramming of
Gene Expression in Amphibian and Insect Cells. New
York: Academic Press.
Gromko, M.H. (1995) Unpredictability of correlated response to selection: Pleiotropy and sampling interact.
Evolution 49:685–693.
Haldane, J.B.S. (1932) The time of action of genes, and
its bearing on some evolutionary problems. Am. Nat.
66:5–24.
Hall, B.K., and S. Hörstadius (1988) The Neural Crest.
New York: Oxford University Press.
Hanken, J., C.H. Summers, and B.K. Hall (1989) Morphological integration in the cranium during anuran
metamorphosis. Experientia 45:872–875.
Hanken, J.H., and D.B. Wake (1993) Miniaturization of
body size: organismal consequences and evolutionary
significance. Annu. Rev. Ecol. Syst. 24:501–519.
Hensel, J.L., Jr., and E.D. Brodie, Jr. (1976) An experimental study of aposematic coloration in the salamander Plethodon jordani. Copeia 1976:59–65.
Hourdry, J., A. L’Hermite, and R. Ferrand (1996)
Changes in the digestive tract and feeding behavior of
anuran amphibians during metamorphosis. Physiol.
Zool. 69:219–251.
Ingham, P.W. (1995) Signalling by hedgehog family proteins in Drosophila and vertebrate development. Curr.
Opin. Genet. Dev. 5:492–498.
Istock, C.A. (1967) The evolution of complex life cycle
phenomena: An ecological perspective. Evolution 21:
592–605.
Johnson, S.L., D. Africa, C. Walker, and J.A. Weston
(1995) Genetic control of adult pigment stripe development in zebrafish. Dev. Biol. 167:27–33.
Kaplan, R.H. (1992) Greater maternal investment can
decrease offspring survival in the frog Bombina orientalis. Ecology 73:280–288.
Kingsolver, J.G., and D.C. Wiernasz (1991) Development, function, and the quantitative genetics of wing
melanin pattern in Pieris butterflies. Evolution 45:
1480–1491.
Kinoshita, T., and F. Sasaki (1994) Adult precursor cells
in the tail epidermis of Xenopus tadpoles. Histochemistry 101:391–396.
Kirkpatrick, M., and D. Lofsvold (1992) Measuring selection and constraint in the evolution of growth. Evolution 46:954–971.
Kusano, T., H. Kusano, and K. Miyashita (1985) Sizerelated cannibalism among larval Hynobius nebulosus. Copeia 1985:472–476.
Lechaire, J.-P., and J.-P. Denefle (1991) Dermal breaks:
Migratory cell pathways opened in anuran tadpole
skin at climactic metamorphosis. Tissue Cell 23:307–
315.
Lehman, H. E. (1953) Analysis of the development of
pigment patterns in larval salamanders, with special
reference to the influence of epidermis and mesoderm.
J. Exp. Zool. 124:571–617.
Mabee, P.M. (1995) Evolution of pigment pattern development in centrarchid fishes. Copeia 1995:586–607.
66
D.M. PARICHY
Maynard Smith, J., R. Burian, S. Kauffman, P. Alberch,
J. Campbell, B. Goodwin, R. Lande, D. Raup, and L.
Wolpert (1985) Developmental constraints and evolution. Q. Rev. Biol. 60:265–287.
McCurdy, H.M. (1971) The succession of cutaneous melanophore populations in Taricha torosa. Can. J. Zool.
49:1075–1078.
Montgomery, J.C., C.F. Baker, and A.G. Carton (1997)
The lateral line can mediate rheotaxis in fish. Nature
389:960–963.
Moran, N.A. (1994) Adaptation and constraint in the
complex life cycles of animals. Annu. Rev. Ecol. Syst.
25:573–600.
Niu, M.C., and V.C. Twitty (1950) The origin of epidermal melanophores during metamorphosis in Triturus
torosus. J. Exp. Zool. 113:633–648.
Noble, G.K. (1931) The Biology of the Amphibia. New
York: Dover Publications, Inc.
Norris, K.S., and C.H. Lowe (1964) An analysis of background color-matching in amphibians and reptiles.
Ecology 45:565–580.
Northcutt, R.G., K.C. Catania, and B.B. Criley (1994)
Development of lateral line organs in the axolotl. J.
Comp. Neurol. 340:480–514.
Olsson, L., and J. Löfberg (1992) Pigment pattern formation in larval ambystomatid salamanders: Ambystoma tigrinum tigrinum. J. Morphol. 211:73–85.
Oster, G.F., N. Shubin, J.D. Murray, and P. Alberch
(1988) Evolution and morphogenetic rules: The shape
of the vertebrate limb in ontogeny and phylogeny.
Evolution 42:862–884.
Parichy, D.M. (1996a) Salamander pigment patterns:
How can they be used to study developmental mechanisms and their evolutionary transformation? Int. J.
Dev. Biol. 40:871–884.
Parichy, D.M. (1996b) Pigment patterns of larval salamanders (Ambystomatidae, Salamandridae): The role
of the lateral line sensory system and the evolution of
pattern-forming mechanisms. Dev. Biol. 175:265–282.
Parichy, D.M. (1996c) When neural crest and placodes
collide: Interactions between melanophores and the
lateral lines that generate stripes in the salamander
Ambystoma tigrinum tigrinum (Ambystomatidae). Dev.
Biol. 175:283–300.
Patterton, D., W.P. Hayes, and Y.B. Shi (1995) Transcriptional activation of the matrix metalloproteinase gene
stromelysin-3 coincides with thyroid-hormone-induced
cell death during frog metamorphosis. Dev. Biol. 167:
252–262
Price, T., and T. Langen (1992) Evolution of correlated
characters. Trends Ecol. Evol. 7:307–310.
Price, T., and M. Pavelka (1996) Evolution of a colour
pattern: History, development, and selection. J. Evol.
Biol. 9:451–470.
Price, T.D., and P.R. Grant (1984) Life history traits and
natural selection for small body size in a population of
Darwin’s finches. Evolution 38:483–494.
Raff, R.A. (1996) The Shape of Life. Chicago, IL: University of Chicago Press.
Ranganayakulu, G., B. Zhao, A. Dokidis, J.D. Molkentin, E.N. Olson, and R.A. Schulz (1995) A series of
mutations in the D-MEF2 transcription factor reveal
multiple functions in larval and adult myogenesis in
Drosophila. Dev. Biol. 171:169–181.
Rauskolb, C., K.M. Smith, M. Peifer, and E. Weischaus
(1995) extradenticle determines segmental identities
throughout Drosophila development. Development
121:3663–3673.
Reedy, M.V., D.M. Parichy, C.A. Erickson, K.A. Mason,
and S.K. Frost-Mason (1998) The regulation of melanoblast migration and differentiation. In J.J. Nordland,
R.E. Boissy, V.J. Hearing, R.A. King, and J.P. Ortonne
(eds): The Pigmentary System and Its Disorders. New
York: Oxford University Press.
Reilly, S.M., and G.V. Lauder (1990) Metamorphosis of
cranial design in tiger salamanders (Ambystoma tigrinum): A morphometric analysis of ontogenetic change.
J. Morphol. 204:121–137.
Rice, W.R. (1989) Analyzing tables of statistical tests.
Evolution 43:223–225.
Riedl, R. (1978) Order in Living Organisms. New York:
Wiley.
Riska, B. (1986) Some models for development, growth,
and morphometric correlation. Evolution 40:1303–
1311.
Riska, B. (1989) Composite traits, selection response,
and evolution. Evolution 43:1172–1191.
Riska, B., and W.R. Atchley (1985) Genetics of growth
predict patterns of brain-size evolution. Science 229:
668–671.
Roach, D.A. (1986) Life history variation in Geranium
carolinianum. I. Covariation between characters at
different stages of the life cycle. Am. Nat. 128:47–57.
Rose, C.S. (1995) Skeletal morphogenesis in the urodele
skull. III. Effect of hormone dosage in TH-induced
remodeling. J. Morphol. 223:243–261.
Rose, C.S., and J.O. Reiss (1993) Metamorphosis and
the vertebrate skull: Ontogenetic patterns and developmental mechanisms. In J. Hanken and B.K. Hall
(eds): The Skull, Vol. 1. Development. Chicago: University of Chicago Press, pp. 289–346.
Selleck, M.A.J., T.Y. Scherson, and M. Bronner-Fraser
(1993) Origins of neural crest cell diversity. Dev. Biol.
159:1–11.
Shaffer, H.B. (1993) Systematics of model organisms:
The laboratory axolotl, Ambystoma mexicanum. Syst.
Biol. 42:508–522.
Shaffer, H.B., and G.V. Lauder (1988) The ontogeny of
functional design: Metamorphosis of feeding behaviour in the tiger salamander (Ambystoma tigrinum). J.
Zool. 216:437–454.
Shaffer, H.B., and M.L. McKnight (1996) The polytypic
species revisited: Genetic differentiation and molecular phylogenetics of the tiger salamander Ambystoma
tigrinum (Amphibia: Caudata) complex. Evolution 50:
417–433.
Shaffer, H.B., and S.R. Voss (1996) Phylogenetic and
mechanistic analysis of a developmentally integrated
character complex: Alternate life history modes in
ambystomatid salamanders. Amer. Zool. 36:24–35.
Shaffer, H.B., C.C. Austin, and R.B. Huey (1991) The
consequences of metamorphosis on salamander (Ambystoma) locomotor performance. Physiol. Zool. 64:212–
231.
Shaw, F.H., R.G. Shaw, G.S. Wilkinson, and M. Turelli
(1995) Changes in genetic variances and covariances:
G whiz! Evolution 49:1260–1267.
Shellenbarger, D.L., and J.D. Mohler (1978) Temperature sensitive periods and autonomy of pleiotropic
effects of l(1)Nts, a conditional Notch lethal in Drosophila. Dev. Biol. 62:432–446.
Shubin, N., D.B. Wake, and A.J. Crawford (1995) Morphological variation in the limbs of Taricha granulosa
(Caudata: Salamandridae): Evolutionary and phylogenetic implications. Evolution 49:874–884.
Slatkin, M. (1987) Quantitative genetics of heterochrony. Evolution 41:799–811.
Smith-Gill, S.J. (1974) Cytophysiological basis of disruptive pigmentary patterns in the leopard frog Rana
pipiens. I. Chromatophore densities and cytophysiology. J. Morphol. 140:271–284.
Sokal, R.R., and F.J. Rohlf (1981) Biometry. New York:
W.H. Freeman and Company.
PIGMENT PATTERN METAMORPHOSIS
Stearner, S.P. (1946) Pigmentation studies in salamanders, with especial reference to the changes at
metamorphosis. Physiol. Zool. 19:375–404.
Stebbins, R.C. (1985) A Field Guide to Western Reptiles
and Amphibians. Boston: Houghton Mifflin Company.
Stine, C., Jr., J.A. Fowler, and R.S. Simmons (1954)
Occurrence of the eastern tiger salamander, Ambystoma tigrinum tigrinum (Green) in Maryland, with
notes on its life history. Ann. Carnegie Mus. 33:145–
148.
Stolow, M.A., and Y.-B. Shi (1995) Xenopus sonic hedgehog as a potential morphogen during embryogenesis
and thyroid hormone-dependent metamorphosis.
Nucleic Acids Res. 23:2555–2562.
Strathman, R.R. (1978) The evolution and loss of feeding
larval stages of marine invertebrates. Evolution 32:
894–906.
Szarski, H. (1957) The origin of the larva and metamorphosis in Amphibia. Am. Nat. 91:283–301.
Taylor, E.H. (1941) Two new ambystomatid salamanders
from Chihuahua. Copeia 1941:143–146.
Twitty, V.C., and D. Bodenstein (1939) Correlated genetic and embryological experiments on Triturus. III:
Further transplantation experiments on pigment development. IV: The study of pigment cell behavior in
vitro. J. Exp. Zool. 81:357–398.
Vaglia, J.L., S.K. Babcock, and R.N. Harris (1997) Tail
development and regeneration through the life cycle
of the four toed salamander Hemidactylium scutatum.
J. Morphol. 233:15–29.
van Eeden, F.J.M., M. Granato, U. Schach, M. Brand, M.
Furutani-Seiki, P. Hafter, M. Hammerschmidt, C.-P.
Heisenberg, Y.-J. Jiang, D.A. Kane, R.N. Kelsh, M.C.
Mullins, J. Odenthal, R.M. Warga, and C. NüssleinVolhard (1996) Genetic analysis of fin formation in the
zebrafish, Danio rerio. Development 123:265–262.
Wagner, G.P., and L. Altenberg (1996) Complex adaptations and the evolution of evolvability. Evolution 50:
967–976.
Wake, D.B. (1991) Homoplasy: The result of natural
selection or evidence of design limitations? Am. Nat.
138:543–567.
67
Wake, D.B., and J.H. Hanken (1996) Direct development
in the lungless salamanders: What are the consequences for developmental biology, evolution, and phylogenesis? Int. J. Dev. Biol. 40:859–869.
Wake, M.H. (1989) Phylogenesis of direct development
and viviparity in vertebrates. In D.B. Wake and G.
Roth (eds): Organismal Functions: Integration and
Evolution in Vertebrates. New York: Wiley. pp 235–
250.
Wassersug, R.J. (1975) The adaptive significance of the
tadpole stage with comments on the maintenance of
complex life cycles in anurans. Am. Zool. 15:405–417.
Wassersug, R.J., and K. Hoff (1982) Developmental
changes in the orientation of the anuran jaw suspension. Evol. Biol. 15:23–246.
Werner, E.E. (1988) Size, scaling, and the evolution of
complex life cycles. In B. Ebenman and L. Persson
(eds): Size-Structured Populations: Ecology and Evolution. Berlin: Springer, pp. 60–81.
Wilbur, H.M. (1980) Complex life cycles. Annu. Rev.
Ecol. Syst. 11:67–93.
Wilder, I.W. (1925) The Morphology of Amphibian Metamorphosis. Northampton, MA: Smith College.
Williamson, D.I. (1982) Larval morphology and diversity. In L.G. Abele (ed): The Biology of Crustacea. New
York: Academic Press, pp. 43–110.
Woronzowa, M.A. (1932) Analyse der wiessen Flecken
bei Amblystomen. Biol. Zentralbl. 52:676–684.
Wray, G.A. (1992) The evolution of larval morphology
during the post-Paleozoic radiation of echinoids. Paleobiology 18:258–287.
Wray, G.A. (1996) Parallel evolution of nonfeeding larvae in echinoids. Syst. Biol. 45:308–322.
Wright, M.R. (1951) The lateral line system of sense
organs. Q. Rev. Biol. 26:264–280.
Wright, S. (1968) Evolution and the Genetics of Populations, Vol. I. Genetic and Biometric Foundations. Chicago: University of Chicago Press.
Yasutomi, M. (1987) Migration of epidermal melanophores to the dermis through the basement membrane during metamorphosis in the frog, Rana japonica. Pigment Cell Res. 1:181–187.